Our long-term goal is to find out how axons in the vertebrate brain determine when and where to form synapses. In particular, we want to identify target-derived signals that guide ingrowing axons to arborize in appropriate laminae, to choose particular postsynaptic partners, and to form functional nerve terminals. Our experimental preparation for this study is the retinotectal projection of the chick, chosen because it is experimentally accessible and well-studied, and because retinotectal connectivity shares features of laminar as well as topographic specificity with the less accessible mammalian cerebral cortex. First, to provide a solid foundation for later molecular analyses, we will describe how retinal axons invade the retinorecipient laminae, form terminal arbors, and differentiate. Similarly, we will assess the appearance of postsynaptic specializations in two well-defined types of tectal neurons that have been identified as direct targets of retinal afferents. Second, we will use immunohistochemical and biochemical methods to identify cell and substrate adhesion molecules that are selectively associated with retinorecipient laminae and thus could play roles in guiding synaptogenesis. We will characterize three such candidates that we have identified so far, and seek others. Third, we will test candidate molecules and mechanisms in a novel retinotectal co- culture system that preserves cues important for lamina-selective outgrowth and arborization. Hypotheses that arise from this work will be tested in vivo, using immunological blockade or virus-mediated gene transfer. Fourth, to extend our analysis to a finer level, we will take advantage of the finding that each of three retinorecipient laminae appears to be innervated by separate subpopulations of retinal ganglion cells. We will define markers for these subpopulations, and use them to study divergent choices that axons make within the retinorecipient zone. Finally, we will extend our work to the mouse superior colliculus, the homologue of the tectum, to ask how retinocollicular innervation is perturbed in mutant animals that lack candidate recognition molecules.